综合成本导向的风电短期功率预测

Comprehensive Cost-oriented Short-term Wind Power Prediction

风电短期功率预测作为电力市场出清等决策问题的输入,其预测精度将直接影响决策效益。传统上风电短期功率预测模型是通过最小化统计误差构建,然而由于正、负预测误差带来的损失成本具有不对称性,预测精度与决策效益存在不匹配的问题。为解决上述问题,该文提出一种综合成本导向的风电短期功率预测方法。该方法直接以最小化新能源预测考核和电力系统运行的综合成本为目标函数训练预测模型。基于内蒙地区风电场群实测数据开展算例验证,结果表明,所提出的预测方法能够有效降低综合成本。

Short-term wind power forecasting,serving as input for decision-making issues such as electricity market clearing,directly impacts decision-making benefits.Traditionally,short-term wind power forecasting models are constructed by minimizing statistical errors.However,due to the asymmetric cost of losses caused by positive and negative forecasting errors,there is a mismatch between forecasting accuracy and decision-making benefits.To bridge the mismatch,this paper proposes a comprehensive cost-oriented short-term wind power prediction method.This method trains forecasting models with the objective of minimizing the comprehensive cost of renewable energy forecasting assessment and power system operation.Case studies based on data from wind farms in the Inner Mongolia region demonstrate that the proposed method can effectively reduce the comprehensive cost.